Prices are effects. We model the causes.
CommonQuant is a thesis-testing engine for markets. You bring a claim about the world. We put it under test and keep it there.
THESIS · EXAMPLE
LIVE
“AI capex is peaking”
VERDICT
CHALLENGED
CONDITIONS
2 of 5 failing
Verdict updates as evidence arrives
ID thesis_0x4a2f
An example thesis under test.
beta
Product stage
9
Checks on every paper order
<1s
Full-market analysis pass
Say you believe something about the market. Then what?
The product has five moving parts. Here is the path a claim takes through them.
01
Thesis
A claim about the market that states what would prove it wrong.
02
Evidence
Data and narratives, each item carrying its source and a trust tier.
03
Test
We try to break the thesis before the market does.
04
Verdict
What we believe now, how confident we are, and what would change our mind.
05
Ledger
Every thesis you have run, including the ones that died.
↺
New evidence sends it back to 02. A thesis here is never finished.
Virtual quant teams
You don't have to work a thesis alone. CommonQuant spawns agent teams that work it the way a quant desk would: one agent researches, another builds candidate strategies, a third tries to tear them down. A thousand of these simulations can run in parallel, on different theses or on different readings of the same one.
The teams compete. Strategies that survive keep evolving, and strategies that fail are archived with their post-mortems. A few months of this leaves you with something sturdier than a model's opinion: strategies that outlived sustained attempts to kill them, with the full paper trail of how each one made it.
Theses also build on each other. A macro thesis about rates can constrain an equity thesis about homebuilders, and a team can inherit both to research the intersection. Your portfolio starts to look less like a list of positions and more like an argument, one you can inspect at any level.
How a strategy earns capital
New strategies start with no money. Here is how one earns it.
01
Simulation
1,000 strategies enter
Strategies are born in closed-loop simulation. No live data leaves the loop and no orders are placed. Most die here.
02
Paper
dozens reach paper
Survivors trade live market data with paper execution. Fills, slippage, and drawdowns are tracked as if the money were real.
03
Seed capital
a few earn real capital
A strategy that holds up on paper earns a small real allocation, behind position limits, loss limits, and a kill switch.
04
Scale
fewer still scale
Allocation grows with demonstrated performance and shrinks the moment behavior drifts from what was validated.
What the ladder really changes is which experiments you can afford. On a live desk, testing an idea risks money, so most ideas are never tested and stay opinions. When the entry rung is simulation, a test costs compute. An idea that merely sounds plausible gets run anyway, in every variation worth trying, and the ones that fail get to fail cheaply, before anyone is attached to them.
What runs today
The product you can use now is the analysis layer — you write theses and the system tests them. The ladder runs deeper inside MarketFabric, our own research and trading engine. Strategy evolution and paper trading are live there, paper since March, with every order passing a nine-check risk gate before it fills. Real capital is the rung we haven't climbed yet, and the paper results decide when we do.